Intensive glycemic control caused a significant reduction in the occurrence of non-fatal myocardial infarctions among patients with type 2 diabetes (T2D) in the ACCORD trial. This beneficial effect, however, was offset by an increase in mortality associated with this intervention. While the reasons for this adverse effect are debated, the task is to devise a treatment strategy by which we can take advantage of the beneficial effects of intensive glycemic control while containing the detrimental ones. To this end, we seek in this ACCORD Ancillary Study to find genetic markers that can identify T2D patients who would especially benefit from intensive glucose-lowering efforts, because of greater sensitivity to the positive effects of this intervention, lesser susceptibility to its adverse effects, or both. Certain clinical characteristics that may help pinpoint these subjects have been identified, but additional predictors are needed to build a robust algorithm. Based on our previous observation of an interaction between degree of glycemic control and the 9p21 CVD locus on the risk of coronary artery disease in T2D, we hypothesize that genetic markers can be used for this task and propose their identification through a systematic search of the entire genome. We propose the following specific aims: 1. To conduct a 733K SNP genome-wide association study (GWAS) to identify genetic modifiers of the effect of intensive glycemic control on cardiovascular outcomes and adverse events in ACCORD. We will test each of the 733,000 loci for interaction with intensive glycemic control on fatal and non-fatal cardiovascular events as well as adverse effects such as severe hypoglycemia and weight gain. We will meta-analyze results with those from ADVANCE through a collaboration with that group. 2. To investigate whether the candidate genetic modifiers identified in ACCORD also influence CVD outcomes in a clinical practice setting. We will study the interaction between these SNPs and long- term glycemic control on cardiovascular outcomes among 2,300 T2D patients from the Joslin Clinic with rich historical HbA1c data. 3. To build prediction models to distinguish T2D patients who are most likely to benefit from intensive glycemic control as compared to standard therapy. We will integrate the clinical and genetic data from ACCORD into regression models and will evaluate their performance in predicting cardiovascular outcomes or adverse events in relation to the type of glucose- lowering therapy. By identifying genetic modulators of the effect of glycemic control on the development of cardiovascular disease, this research will provide a starting point to build a personalized medicine framework to treat T2D patients in a more cost-effective way. Identification of these genetic factors may also provide novel insights into the molecular pathways linking excess glucose to atherosclerosis, with critical implications for the development of novel drugs to prevent CVD in T2D.